{"id":8253,"date":"2025-02-09T06:26:40","date_gmt":"2025-02-09T06:26:40","guid":{"rendered":"https:\/\/howtogeek.blog\/lv\/?p=8253"},"modified":"2025-02-09T06:26:40","modified_gmt":"2025-02-09T06:26:40","slug":"paris-ai-summit-experts-call-for-stricter-regulations-to-maintain-control-over-ai-development","status":"publish","type":"post","link":"https:\/\/howtogeek.blog\/lv\/paris-ai-summit-experts-call-for-stricter-regulations-to-maintain-control-over-ai-development\/","title":{"rendered":"Par\u012bzes AI samits: eksperti aicina ieviest stingr\u0101kus noteikumus, lai saglab\u0101tu kontroli p\u0101r AI att\u012bst\u012bbu"},"content":{"rendered":"

Glob\u0101lajiem l\u012bderiem Par\u012bz\u0113 pulc\u0113joties galvenajam samitam, kur\u0101 galven\u0101 uzman\u012bba piev\u0113rsta m\u0101ksl\u012bgajam intelektam, da\u017e\u0101du nozaru eksperti iest\u0101jas par nepiecie\u0161am\u012bbu p\u0113c stingriem m\u0101ksl\u012bg\u0101 intelekta noteikumiem, lai saglab\u0101tu \u0161\u012bs sp\u0113c\u012bg\u0101s tehnolo\u0123ijas cilv\u0113ka uzraudz\u012bbu.<\/p>\n

\u0160\u012b gada pulc\u0113\u0161an\u0101s, ko kop\u012bgi r\u012bko Francija un Indija, iez\u012bm\u0113 p\u0101reju no iepriek\u0161\u0113jiem notikumiem, piem\u0113ram, pulc\u0113\u0161an\u0101s Bletchley park\u0101 un Seul\u0101, kur galvenok\u0101rt tika piev\u0113rsta uzman\u012bba AI dro\u0161\u012bbai.2024.gada samita m\u0113r\u0137is ir ietvert pla\u0161\u0101ku skat\u012bjumu uz AI p\u0101rvald\u012bbu, ilgtsp\u0113j\u012bbu un nozares atbild\u012bbu.<\/p>\n

P\u0101reja uz AI p\u0101rvald\u012bbu un iesp\u0113j\u0101m<\/strong><\/span><\/h2>\n

Francijas darba k\u0101rt\u012bba apzin\u0101ti uzsver ne tikai riskus, bet ar\u012b AI sniegt\u0101s iesp\u0113jas.\u201cM\u0113s nev\u0113lamies run\u0101t tikai par riskiem.\u0160eit ir ar\u012b \u013coti re\u0101lu iesp\u0113ju aspekts, \u201d<\/strong> sac\u012bja Anne Buvero, Francijas prezidenta Emanuela Makrona AI s\u016btne.<\/p>\n

Tom\u0113r ap \u0161o optimistisko skat\u012bjumu valda skepse. Dz\u012bv\u012bbas n\u0101kotnes instit\u016bta prezidents Makss Tegmarks br\u012bdina, ka, lai gan iesp\u0113jas past\u0101v, t\u0101s nedr\u012bkst aiz\u0113not steidzamas darb\u012bbas pret pieaugo\u0161ajiem AI rad\u012btajiem draudiem.“Francija ir bijusi br\u012bni\u0161\u0137\u012bga starptautisk\u0101s sadarb\u012bbas \u010dempione, un tagad tai ir iesp\u0113ja vad\u012bt glob\u0101lo sabiedr\u012bbu, “<\/strong> koment\u0113ja iev\u0113rojamais MIT fizi\u0137is.<\/p>\n

AI risku nov\u0113r\u0161ana ar glob\u0101l\u0101m iniciat\u012bv\u0101m<\/strong><\/span><\/h2>\n

Proakt\u012bvi samits ir ieviesis glob\u0101l\u0101s riska un AI dro\u0161\u012bbas sagatavot\u012bbas (GRASP) sist\u0113mu.\u0160\u012b iniciat\u012bva, ko atbalsta Future of Life instit\u016bts, cen\u0161as analiz\u0113t un mazin\u0101t ar AI saist\u012btos riskus, nosakot efekt\u012bvus glob\u0101l\u0101s dro\u0161\u012bbas risin\u0101jumus.“M\u0113s esam identific\u0113ju\u0161i aptuveni 300 r\u012bkus un tehnolo\u0123ijas, lai nov\u0113rstu \u0161os riskus, “<\/strong> paskaidroja GRASP koordinators Sairuss Hodess.<\/p>\n

Nesen public\u0113taj\u0101 revolucion\u0101raj\u0101 starptautiskaj\u0101 m\u0101ksl\u012bg\u0101 intelekta dro\u0161\u012bbas zi\u0146ojum\u0101 \u2014 96 ekspertu atzi\u0146u apkopojums, ko apstiprin\u0101ju\u0161as valstis, tostarp 30 valstis, k\u0101 ar\u012b lielas organiz\u0101cijas, piem\u0113ram, ANO un ESAO, ir uzsv\u0113rti da\u017e\u0101di riski, s\u0101kot no dezinform\u0101cijas l\u012bdz nopietniem draudiem, piem\u0113ram, biolo\u0123iskiem uzbrukumiem un kiberkaram.<\/p>\n

Iev\u0113rojams m\u0101ksl\u012bg\u0101 intelekta p\u0113tnieks un Tj\u016bringa balvas sa\u0146\u0113m\u0113js Jo\u0161ua Bengio pauda ba\u017eas par risku zaud\u0113t kontroli p\u0101r AI tehnolo\u0123ij\u0101m, apgalvojot, ka “pier\u0101d\u012bjumi past\u0101v\u012bgi par\u0101d\u0101s papildu riskiem, un ilgtermi\u0146\u0101 AI var att\u012bst\u012bt savus izdz\u012bvo\u0161anas instinktus, izraisot neparedz\u0113tas sekas.”<\/strong><\/p>\n

Paredzamie m\u0101ksl\u012bg\u0101 visp\u0101r\u0113j\u0101 intelekta draudi<\/strong><\/span><\/h2>\n

Viena no galvenaj\u0101m AI speci\u0101listu ba\u017e\u0101m ir strauja virz\u012bba uz m\u0101ksl\u012bgo visp\u0101r\u0113jo intelektu (AGI), kas sp\u0113j p\u0101rsp\u0113t cilv\u0113ka intelektu da\u017e\u0101d\u0101s jom\u0101s. Sems Altmans, OpenAI izpilddirektors, kop\u0101 ar Dario Amodei no Anthropic ir ierosin\u0101ju\u0161i, ka AGI iera\u0161an\u0101s var\u0113tu sakrist ar n\u0101kamajiem gadiem.<\/p>\n

“Ja paskat\u0101s tikai uz progresa tempu, tas liek dom\u0101t, ka m\u0113s to sasniegsim l\u012bdz 2026.vai 2027.gadam, ” novembr\u012b atz\u012bm\u0113ja Amodei. Tom\u0113r Tegmark izdeva stingru br\u012bdin\u0101jumu par iesp\u0113jamo scen\u0101riju, kur\u0101 AI l\u012bderi, \u012bpa\u0161i ASV un \u0136\u012bn\u0101, zaud\u0113s kontroli p\u0101r AGI, kas, iesp\u0113jams, noved\u012bs pie pasaules, kur\u0101 domin\u0113 vied\u0101s ma\u0161\u012bnas.<\/p>\n

AI kar\u0101: pieaugo\u0161as ba\u017eas<\/strong><\/span><\/h2>\n

Viens no noz\u012bm\u012bg\u0101kajiem t\u016bl\u012bt\u0113jiem draudiem ir autonomo iero\u010du jom\u0101, ko darbina AI. Stjuarts Rasels, UC Berkeley profesors un Starptautisk\u0101s dro\u0161as un \u0113tiskas AI asoci\u0101cijas (IASEI) koordinators, uzsv\u0113ra savas ba\u017eas par AI sist\u0113m\u0101m, kas pie\u0146em kaujas lauka l\u0113mumus.“Liel\u0101k\u0101s bailes rada AI kaujas lauka l\u0113mumu pie\u0146em\u0161anu \u2014 izv\u0113loties, kam un kad uzbrukt \u2014 bez cilv\u0113ka uzraudz\u012bbas, “<\/strong> vi\u0146\u0161 nor\u0101d\u012bja.<\/p>\n

Eksperti vienbals\u012bgi apgalvo, ka vald\u012bb\u0101m ir nekav\u0113joties j\u0101r\u012bkojas, lai regul\u0113tu m\u0101ksl\u012bgo intelektu ar t\u0101du pa\u0161u stingr\u012bbu k\u0101 citas augsta riska nozares.“Ja k\u0101ds v\u0113las b\u016bv\u0113t kodolreaktoru, vi\u0146am pirms b\u016bvniec\u012bbas j\u0101pier\u0101da, ka tas ir dro\u0161s. AI ir j\u0101izturas t\u0101pat, “<\/strong> uzsv\u0113ra Tegmarks.<\/p>\n

Par\u012bz\u0113 risinot diskusijas, aktu\u0101ls paliek jaut\u0101jums: k\u0101 pasaule var izmantot milz\u012bgo AI potenci\u0101lu, vienlaikus nodro\u0161inot, ka tas neizvair\u0101s no cilv\u0113ka kontroles? N\u0101kamie m\u0113ne\u0161i ir iz\u0161\u0137iro\u0161i, lai noteiktu, vai efekt\u012bva AI p\u0101rvald\u012bba var iet kopsol\u012b ar straujo tehnolo\u0123iju att\u012bst\u012bbu, vai ar\u012b cilv\u0113ce risk\u0113s atpalikt.<\/p>\n

Bie\u017ei uzdotie jaut\u0101jumi<\/h2>\n

1. K\u0101p\u0113c Par\u012bzes samits par AI ir noz\u012bm\u012bgs?<\/strong><\/h3>\n

Par\u012bzes samits ir \u013coti svar\u012bgs, jo taj\u0101 tiek risin\u0101ts gan m\u0101ksl\u012bg\u0101 intelekta potenci\u0101ls, gan riski, galveno uzman\u012bbu piev\u0113r\u0161ot starptautisku noteikumu izveidei, lai nodro\u0161in\u0101tu, ka m\u0101ksl\u012bgais intelekts joproj\u0101m ir cilv\u0113ka kontrol\u0113, vienlaikus veicinot novatoriskas p\u0101rvald\u012bbas strat\u0113\u0123ijas.<\/p>\n

2. K\u0101das iniciat\u012bvas tika uzs\u0101ktas, lai nov\u0113rstu AI dro\u0161\u012bbas riskus?<\/strong><\/h3>\n

Global Risk and AI Safety Preparedness (GRASP) platforma tika ieviesta, lai analiz\u0113tu un mazin\u0101tu ar AI tehnolo\u0123iju saist\u012btos riskus. T\u0101s m\u0113r\u0137is ir identific\u0113t aptuveni 300 r\u012bkus un strat\u0113\u0123ijas, lai efekt\u012bvi c\u012bn\u012btos pret iesp\u0113jamiem draudiem.<\/p>\n

3. K\u0101das ir ba\u017eas saist\u012bb\u0101 ar m\u0101ksl\u012bgo visp\u0101r\u0113jo intelektu (AGI)?<\/strong><\/h3>\n

Eksperti pau\u017e ba\u017eas, ka AGI var p\u0101rsp\u0113t cilv\u0113ka intelektu, izraisot scen\u0101rijus, kad AI sist\u0113mas darbojas autonomi, kas var izrais\u012bt neparedz\u0113tas sekas un kontroles p\u0101r kritiskaj\u0101m sist\u0113m\u0101m zaud\u0113\u0161anu.<\/p>\n

Avots un att\u0113li<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Glob\u0101lajiem l\u012bderiem Par\u012bz\u0113 pulc\u0113joties galvenajam samitam, kur\u0101 galven\u0101 uzman\u012bba piev\u0113rsta m\u0101ksl\u012bgajam intelektam, da\u017e\u0101du nozaru eksperti iest\u0101jas par nepiecie\u0161am\u012bbu p\u0113c stingriem m\u0101ksl\u012bg\u0101 intelekta noteikumiem, lai saglab\u0101tu \u0161\u012bs sp\u0113c\u012bg\u0101s tehnolo\u0123ijas cilv\u0113ka uzraudz\u012bbu. \u0160\u012b gada pulc\u0113\u0161an\u0101s, ko kop\u012bgi r\u012bko Francija un Indija, iez\u012bm\u0113 p\u0101reju no iepriek\u0161\u0113jiem notikumiem, piem\u0113ram, pulc\u0113\u0161an\u0101s Bletchley park\u0101 un Seul\u0101, kur galvenok\u0101rt tika piev\u0113rsta uzman\u012bba […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[335],"class_list":["post-8253","post","type-post","status-publish","format-standard","hentry","category-how-to","tag-news"],"acf":[],"_links":{"self":[{"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/posts\/8253","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/comments?post=8253"}],"version-history":[{"count":1,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/posts\/8253\/revisions"}],"predecessor-version":[{"id":8254,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/posts\/8253\/revisions\/8254"}],"wp:attachment":[{"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/media?parent=8253"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/categories?post=8253"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/howtogeek.blog\/lv\/wp-json\/wp\/v2\/tags?post=8253"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}